ScreenshotToCode: FREE Text-to-Frontend AI with 100% LOCAL Setup (OpenUI & v0 Alternative)

[@lcarv20]

Hi there, I came across your channel recently. Great videos! I am gonna try this definitely. Keep the good work!

[@swapwarick]

I am getting error generating the code. Check developer console and backend logs for details.

[@AICodeKing]

Try also changing the 10th or 11th line to refer to the Lava model. It should fix it.

[@swapwarick]

@@AICodeKing ok I'll try that. But I already did change the lines. Maybe I'll try again.

[@xtro7]

Very interesting, wich would be your recomendation for sending a promt to front? Like controling IU with your voice, for example...

[@bladethirst1]

This seems to be only static content, is there some tool for webpage with reactive component?

[@regallux6973]

Can I use the code with bolt

[@PeterBLiu]

thanks, the frontend works, but after uploading a screenshot, it says "Invalid model: llava"

[@AICodeKing]

Try also changing the 10th or the the 11th line in the file to also point to lava model. That should fix it.

[@PeterBLiu]

@@AICodeKing already changed those two lines as you said, but got the same error, how to fix?

[@fairyroot1653]

I tried everything and nothing worked, I abandoned trying further for it didn't work for you as well, that's why you switched to GPT 4 turbo

[@dtory]

Can you try with gpt 4 or gpt4o

[@dtory]

Also can you include the GitHub repo 😊

[@jackflash6377]

quick search for "screenshot to code" will get you the repo address.

[@redaelhili9254]

Great video as always

[@AICodeKing]

Weird! Works for me though. Maybe, try changing the 10th or 11th line to also refer to Llava. Maybe that's making an issue. Mine works fine without changing them though.

[@johnaroj]

@@AICodeKing same

[@connectedquest]

same here, stuck at this error "Invalid model: llava" 😕

[@connectedquest]

In your video you are not using llama to generate the UI code, as it shows GPT-4 turbo :S

[@vertigoz]

in my case it doesn't run, it throws me an error: "gpt-4o-2024-05-13": { name: "GPT-4o 🌟", inBeta: false }, pointing to gpt-4o-2024-05-13

[@AICodeKing]

Try also changing the 10th or the the 11th line in the file to also point to lava model. That should fix it.

[@Michael-ci1gf]

@@AICodeKing it gets it working but unfortunately throws an error after when you press "update" for the prompt

[@fumbling_forward]

Which GPU are You usinng in lightningAI platform

[@AICodeKing]

T4

[@freeideas]

The problem I am having, is that I don't know how to make attractive front-ends. Coding them is no problem. Making them look good is super difficult for me. I wish there was an Ai that I could describe what controls I want, and the purpose of the screen, and it gives me a screenshot that is visually attractive. Then I could use this tool to write code from the screenshot.

[@freeideas]

I have tried Gpt4o and other image-generation AIs, but they don't seem to want to make mockup screenshots. They will show me a computer sitting on a desk, or a hand holding a cellphone. Maybe my prompting skills are to blame.

[@k2an]

awesome tuts really, but this is not working, i think anymore :) same error , Invalid model: llava but nice apps that i dont know. Congrats!

[@warlockassim4240]

I like gpt4 to do this